Multi-modal Coupling Stability Prediction And Time-frequency Transform Chatter Detection For Thin-walled Workpiece Milling

Because of its light weight and high specific strength,thin-walled parts are widely used in many fields such as aerospace,automobile,mold,energy,rail transit and many other fields.Thin-walled parts have four typical structural characteristics,such as thin-walled weak rigidity,variable cross-section thickness,bending and sweeping characteristics and irregular development characteristics,as well as four cutting characteristics,such as boundary characteristics,time-varying characteristics,modal coupling characteristics and position-dependent characteristics.Therefore,severe vibration or even chatter will occur during the cutting process of thin-walled parts,which will seriously affect the quality of the cutting surface,shorten the tool life and reduce the processing efficiency.In this thesis,the multi-modal coupling stability prediction and time-frequency transform chatter detection of thin-walled parts milling system are studied in depth.A modal coupling method for predicting the milling stability of thin-walled parts with modal coupling effect is proposed.The chatter detection of milling acoustic signals by using cmor continuous wavelet transform and parametric time-frequency transform is proposed.The accurate stability prediction,efficient on-line chatter detection and precise chatter detection of the position-dependent characteristic system for thin-walled parts are realized,which provides the precondition for parameter optimization and processing stability control of high-efficiency and high-precision thin-walled parts.Firstly,aiming at the multi-modal coupling characteristics of thin-walled parts milling system,a prediction method of modal coupling stability is proposed in this paper.Based on the mechanism of multi-modal coupling effect of thin-walled parts,the coupling law of the first several modes of thin-walled parts is discussed,the position-dependent dynamic model of thin-walled parts milling is established,and a prediction method for milling stability of thin-walled parts with position-dependent characteristics and modal coupling effect is proposed.Compared with single mode method and lowest envelope method,the advantages of modal coupling method in predicting the milling stability of thin-walled parts are proved.The validity and accuracy of the modal coupling method are verified by milling surface topography analysis.By means of modal coupling method,the stability lobe diagrams of multi-modal coupling milling can be obtained accurately,and the processing parameters can be optimized.The material removal rate can be maximized while avoiding chattering,and the high efficiency and high precision milling of thin-walled parts can be realized.Secondly,in order to meet the requirement of efficient on-line chatter detection in thin-walled parts milling,an efficient chatter detection scheme based on cmor continuous wavelet transform(CMWT)is proposed in this paper.The monitoring time is only 0.12s and the monitoring accuracy is high.Based on the position-dependent characteristic mechanism of thin-walled parts milling,the time-frequency transform analysis mechanism of milling acoustic signals is discussed,and a cmor continuous wavelet transform(CMWT)scheme for efficient on-line chatter monitoring in thin-walled parts milling is formed.By comparing the chatter detection results of CMWT with that of short-time Fourier transform,the superiority of CMWT in the chatter detection of original acoustic signals in thin-walled parts milling is proved.The accuracy of CMWT in monitoring chatter of thin-walled plate milling is verified by analyzing the stability lobe diagram and milling surface topography.CMWT chatter detection scheme can realize multi-scale analysis and monitoring of non-stationary milling signals.Its monitoring time is only 0.12s and the monitoring accuracy is high,which lays a solid foundation for the follow-up on-line chatter detection.Finally,in order to meet the requirements of time-frequency energy concentration for chatter detection of thin-walled parts,a precise chatter detection scheme based on Spline Chirplet Transform(SCT)and Generalized Warblet Transform(GWT)is proposed in this paper.The transformation kernel function method based on SCT achieves the approximation of milling acoustic signal by spline transformation kernel function,which makes SCT obtain excellent time-frequency performance of energy concentration,and avoids the pathological problem of high-order polynomial fitting and Runge phenomenon.The transform kernel parameter estimation method based on GWT adopts Fourier series kernel and effectively and adaptively determines the coefficients of the Fourier series core based on milling acoustic signal.It achieves better instantaneous frequency approximation between the kernel function and acoustic signal,and achieves excellent time-frequency performance of energy concentration for GWT.By comparing the chatter detection results of SCT,GWT and CMWT,the excellent time-frequency performance and superiority of SCT and GWT in the chatter detection process of strong time-varying non-stationary milling acoustic signals are verified.The accuracy of SCT and GWT chatter detection is verified by milling surface topography analysis.Both SCT and GWT methods have excellent time-frequency energy concentration performance in the chatter detection of strong time-varying non-stationary milling acoustic signals.They can identify the characteristics of milling mode concentration.Moreover,SCT can also identify the position-dependent characteristics of modes in the milling process of thin-walled parts,which is suitable for accurate chatter detection of thin-walled parts in milling.